Saliency Detection with Recurrent Fully Convolutional Networks

Linzhao Wang,Lijun Wang,Huchuan Lu,Pingping Zhang,Xiang Ruan
DOI: https://doi.org/10.1007/978-3-319-46493-0_50
2016-01-01
Abstract:Deep networks have been proved to encode high level semantic features and delivered superior performance in saliency detection. In this paper, we go one step further by developing a new saliency model using recurrent fully convolutional networks (RFCNs). Compared with existing deep network based methods, the proposed network is able to incorporate saliency prior knowledge for more accurate inference. In addition, the recurrent architecture enables our method to automatically learn to refine the saliency map by correcting its previous errors. To train such a network with numerous parameters, we propose a pre-training strategy using semantic segmentation data, which simultaneously leverages the strong supervision of segmentation tasks for better training and enables the network to capture generic representations of objects for saliency detection. Through extensive experimental evaluations, we demonstrate that the proposed method compares favorably against state-of-the-art approaches, and that the proposed recurrent deep model as well as the pre-training method can significantly improve performance.
What problem does this paper attempt to address?